The Project 801 deliverable, DIPPR® 801 Database, contains recommendation values and estimates of accuracy for 34 constant properties and 15 temperature-dependent properties for compounds of interest to entities involved in manipulating chemicals for new products and manufacturing processes. More recently DIPPR has been adding selected EH&S data (e.g. Henry’s Law constants, Auto-ignition temperature) to broaden its value to end-users.
DIPPR differentiates itself from other databases in the thermophysical property fields by subjecting the database to an extensive GOLD STANDARD evaluation methodology which provides insights into the currency and quality of the database as well as providing guidance for our experimental work as to data required to address data gaps, improve data considered of poor quality, or would help improve our proprietary estimation methods.
The DIPPR process uses a systems approach in evaluating data from all available sources to triangulate on the best values, resulting in higher accuracy than could be achieved using only individual points. This approach includes hundreds of applied constraints on property values stemming from inter-property relationships, expected trends of properties between related chemicals, and the impact of chemical similarities and differences. Once these constraints are simultaneously satisfied for all properties, experts review the compound and give final approval for the compound to be added to the database. Each compound in the database therefore has recommended values for all properties and comments as to which of all reported property values is most reliable. When experimental data are not available, values for properties are predicted using both established methods as well as in-house prediction methodologies, and these predicted properties are subject to the same rigorous systems-level evaluation process.
Hallmark 1: The most accurate recommended values and temperature-dependent correlations available.
- Marker 1: Raw data – The database includes a comprehensive collection of experimental raw data for chemicals included in the database. This is primarily a Process Marker and is codified in DIPPR’s Policies and Procedures Manual ©.
- Marker 2: Database currency –Evaluated using an in-house derived weighting factor. Marker 2 is measureable in terms of a Impact Factor (IF) which is an indicator of the impact that “unevaluated” data could have on the recommended constant property values and the temperature-dependent correlations. Whenever a new property value is reported in the literature, the IF indicates when a property is out of date and a “matrix” review is warranted.
- Marker 3: Accuracy – The recommended values and temperature correlations are based on a critical, comprehensive evaluation by experts of the available data including inter-property relationships, inter-chemical relationships, and predicted/expected values. This includes evaluation by one or more industrial sponsors. Marker 3 is mainly a process standard. Every new chemical, every systematic-review chemical, and every matrix-review chemical undergoes the same rigorous evaluation by senior staff, the project coordinator, and the sponsor review panel. This process standard is unique to the DIPPR project and is the heart of the Gold Standard. Additionally, 58 automated QC checks are used in the process to eliminate potential errors.
Hallmark 2: A complete set of properties for each chemical in the database.
- Marker 4: Unmeasured properties – The most accurate prediction methods available have been used to predict properties that have not been measured. These methods include in-house evaluation based on inter-property and inter-chemical relationships. Marker 4 is also a process marker, primarily satisfied by adherence to the DIPPR Policies and Procedures Manual ©.
- Marker 5: Unknown and inappropriate properties – Properties that do not apply to certain chemicals have appropriate notes documenting decomposition, sublimation, and other chemical and physical constraints. Marker 5 is the main indicator of the “completeness” hallmark of the DIPPR database.
Godwin TongoChair, Technical Committee BP
Ashok DewanChair, Technical Committee Shell Global Solutions (US), Inc.
Project Technical LeadPaul Mathias Project 801 Technical Lead Fluor Corporation
Alvin H. Larsen
Solutia Inc., St. Louis, MO
The Design Institute for Physical Properties (DIPPR®) of the American Institute of Chemical Engineers has become the premier cooperative physical property data effort in the United States to satisfy process engineering needs. Celebrating its 28th Anniversary in November 2006, DIPPR® has been active longer than any other sponsored research group of AIChE. The background and formation of DIPPR® are described, highlighting the people responsible for launching DIPPR® on its path of ongoing success.
Early Data Efforts: Cooperative data efforts in the United States were carried out by a wide variety of organizations in the 1960s and 1970s. Government agencies such as the National Bureau of Standards (NBS, later the National Institute for Standards and Technology, NIST), the Energy Research and Development Administration, and the Environmental Protection Agency conducted projects primarily with federal funding. Industry associations, including the American Petroleum Institute (API), the Manufacturing Chemists Association (MCA, later the Chemical Manufacturers Association, CMA), the Natural Gas Producers Association, Fluid Properties Research, Inc. (FPRI), and the Energy and Power Research Institute, were directed and funded by participating companies. Universities such as the Massachusetts Institute of Technology, the University of Houston, and Purdue University developed data programs primarily with federal grants, and the Thermodynamics Research Laboratory (TRL) at Washington University was supported both by federal grants and industry participants. Simulation companies such as ChemShare and Simulation Sciences developed data banks for their software products.
Two long-standing projects served as examples of compilations of critically evaluated data. The API Research Project 44, which began at NBS in 1942 and moved to the Carnegie Institute of Technology (later the Carnegie-Mellon University) in 1950, developed and published reliable property data for hydrocarbons. The MCA Research Project began as a companion project for non- hydrocarbons at the Carnegie Institute of Technology in 1955. Both these projects were moved to the Thermodynamics Research Center (TRC) at Texas A&M University in 1961 and were renamed.
Other efforts around the world included the Physical Property Data Service (PPDS) of the Institution of Chemical Engineers (U.K.); the program of the Engineering Sciences Data Unit (ESDU, later ESDU International Ltd.), supported by the National Physical Laboratory, the National Engineering Laboratory, and other British technical organizations; the Uhde physical property program, marketed by DECHEMA, the German Chemical Engineering Association (Germany); the JUSE-AESOPP physical properties program, from the Japanese Union of Scientists and Engineers (Japan); and the program of the CODATA Task Group on Data for Industrial Chemicals (International Council of Scientific Unions). Project Evergreen
In preparing the third edition of "The Properties of Gases and Liquids" in 1974, Bob Reid of the Massachusetts Institute of Technology wanted to have a data bank of important compounds included as an appendix. He asked Union Carbide and, later, several other companies to contribute their data. He compared values and found many points of agreement and many errors. As a result, Don Vredeveld of Union Carbide suggested further cooperative efforts in physical property data.
Bob Reid then organized a meeting November 18, 1975, at the AIChE Annual Meeting in Los Angeles to determine the level of interest in a cooperative data effort. Over 30 people attended. Don Vredeveld suggested sharing a data bank and maintaining it in "evergreen" fashion. Howard White of the NBS Office of Standard Reference Data (OSRD) developed a proposal for "Project Evergreen" and organized a meeting at NBS in Gaithersburg, MD, September 21, 1976, to discuss it further and determine support for it. No strong objections were raised, but enthusiasm for the proposal was lukewarm. The Project Evergreen proposal was refined and presented to about 50 people at a meeting organized by the AIChE Machine Computation Committee at the 1976 Annual Meeting, November 30, in Chicago.
Project Evergreen envisioned a data base of about a thousand "ubiquitous" compounds, including all the properties required for process engineering. The data base would be created with the best available data and estimates, and made "evergreen" by continually updating it as better data and methods were found. No experimental programs were contemplated, and maximum use would be made of existing data compilations. It was proposed that responsibility for the data base and the data center would be assumed jointly by industry and the federal government, and that the data center would be under the program management of OSRD. A five-year program was proposed, with a budget of at least $100,000 per year. A wide diversity of opinions was expressed, on issues such as scope, funding, government control, quality of results, and response to industry needs. Additional discussion was held at the first international conference on Fluid Properties and Phase Equilibria in the Chemical Industry, at Asilomar, CA, in mid-January 1977.
MCA Initiative: To handle funding and legal issues, an umbrella organization for industry participation in such a cooperative effort was considered essential. Several potential umbrella organizations were discussed. MCA, AIChE, NBS, and API were suggested. MCA was considered first, since its membership consisted of a wide variety of companies in the chemical industries and since it had previously coordinated a data project at Carnegie Institute of Technology and, later, at Texas A&M University.
At a meeting in Washington October 27, 1976, the MCA Engineering Advisory Committee appointed David Roth of Allied Chemical as chairman of a Thermophysical and Chemical Data Ad Hoc Task Group. The Task Group was established to review the Project Evergreen proposal, evaluate other alternatives, and recommend the level of MCA involvement in such an endeavor.
David Roth called a meeting of the Task Group, held January 10, 1977, at Allied Chemical's offices in Morristown, NJ. The Task Group concluded that Project Evergreen had several drawbacks which could be overcome by MCA sponsorship of the project. It later recommended that MCA establish a permanent committee to oversee the effort. The Task Group was superseded in April 1977 by a Subcommittee on Thermodynamic and Physical Data under the Engineering Advisory Committee, with David Roth as chairman.
Several discussions and meetings were held in Washington under MCA auspices during 1977. The program objectives, contractor selection, and data base content were developed in some detail. Mel Albright of Phillips Petroleum suggested an organization similar to the Gas Processors Association (GPA) and a project funding procedure patterned after that successfully used by GPA, in which projects were funded separately by annual commitments of the sponsoring companies through a balloting procedure. This approach was adopted, and Evan Buck of Union Carbide developed additional details of an organizational structure and a funding mechanism. David Palmer of Amoco Chemicals drafted a program proposal and developed a proposal for a project to measure acetic acid-water PVT and VLE data, and Al Larsen of Monsanto developed a proposal for a data base project (later the Data Compilation project) and organized the program proposal for the entire effort, including the contributions by Evan Buck and David Palmer.
Concerns regarding MCA sponsorship were raised, including use of the MCA "Special Project" format for an open-ended project, participation of non-MCA members in technical direction and funding, administrative procedures, and possible use of OSRD as the program manager with technical direction and funding through MCA. The MCA staff reviewed the proposal in March 1978, and would not recommend approval to the MCA Board of Directors. The Subcommittee under David Roth then recommended that the proposal be withdrawn from MCA consideration.
AIChE Sponsorship: Over the same period of time, AIChE was also considered as a potential sponsor of the cooperative physical property data effort. Steve Newman of Foster Wheeler suggested in February 1977 that the AIChE Research Committee be involved, and later proposed a project for preparation of a data book similar to the API Technical Data Book ¾ Petroleum Refining, but oriented toward chemicals. In September 1977, in Chicago, David Palmer discussed the MCA proposal with AIChE leadership, to explore possibilities of AIChE sponsorship.
In October 1977, David Roth requested that an acronym be selected for a proposal to AIChE Council the following month. Over 25 acronyms were considered for the effort with MCA, but none of them had strong support. Of course, the acronym would be derived from and be consistent with the name of the effort. But it also had to be catchy and easy to remember. In response to the request, Al Larsen suggested "DIPPR," from its organization as a Design Institute of AIChE and from the need to include Physical Property Data (or some close equivalent) in a direct way. The emphasis was to be on Data, not Research, but "DIPPD" just didn't have a good ring to it. "DIPPR" could be used anyway, by taking the "R" from "PRoperty." The acronym has served well and is highly recognizable.
The proposal was presented to AIChE Council in New York in November 1977 to establish DIPPR with an organizational structure similar to that of the Design Institute for Emergency Relief Systems (DIERS), which had recently begun. John Prados of the University of Tennessee, a former AIChE Director, was appointed to head an ad hoc committee to study the proposal.
It was evident that AIChE would be very interested in serving as the umbrella organization for the proposed program, as a Design Institute, should the effort with MCA not come to fruition. In January 1978, Stan Adler of Pullman Kellogg, Al Larsen, Steve Newman, David Palmer, David Roth, and F. J. Van Antwerpen, Executive Director of AIChE, were also named to the ad hoc committee, which deferred action pending the MCA decision. Word was finally received in late April 1978 that the MCA initiative was withdrawn. The ad hoc DIPPR® committee met a few days later in New York and moved rapidly to revise the proposal for presentation to AIChE Council in Philadelphia June 3. At that time, Council approved formation of DIPPR®, disbanded the ad hoc committee, and asked John Prados to chair a DIPPR® Formation Committee. He named Al Larsen, Steve Newman, David Palmer and David Roth to the committee.
John Prados established a timetable to launch the DIPPR® organization, including publicity and solicitation for membership, an organizational meeting at the AIChE Annual Meeting in November 1978, balloting for 1980 projects in early 1979, selection of project investigators in Fall 1979, and beginning the second-year cycle at the 1979 Annual AIChE Meeting. AIChE headquarters sent out over 600 letters of solicitation for DIPPR participation over the signature of Wm. H. Corcoran, AIChE President.
During the three years it took to get the cooperative physical property data effort organized as DIPPR®, under AIChE auspices, the proposal was refined and expanded several times. Momentum for the effort increased with each meeting. Meetings were held in many locations, throughout the country, converging on the organizational meeting in Miami Beach.
Organizational Meeting: The long-awaited meeting was convened in the Americana Bal Harbour Hotel in Miami Beach, November 15, 1978, at 9:00 AM. A sense of excitement and anticipation filled the air. It was also the first meeting of the DIPPR® Technical Committee. The meeting was held in two parts: (1) an open session, for all interested people, and (2) a closed session, for representatives of DIPPR® participants only. DIPPR® participants were defined as those companies who had paid the $500 DIPPR® dues for 1979, but due to the short time available after the solicitation letters were sent until the organizational meeting, companies whose representatives were confident of joining DIPPR® were also invited to the closed session.
The open session introduced DIPPR, including objectives, organization, funding methods, schedules, and proposed projects. The closed session discussed weighting factors for various types of companies, and elected John Prados Chairman of the Technical Committee, with David Palmer as Vice Chairman. Project steering committees for proposed projects, as well as the Liaison Committee and the Publicity and Recruitment Committee, were formed by electing chairmen of the committees. These chairmen were then made responsible to recruit their committee members from interested Technical Representatives. Meetings of the Technical Committee were set to be held annually, in conjunction with AIChE Annual Meetings.
Five projects were approved for balloting to DIPPR® participants: Data Compilation, Chemicals Technical Data Book, Electrolyte Phase Equilibria, Acetic Acid-Water PVT and VLE Measurements, and Mixture Experimental Data. These projects were balloted in early 1979 to begin work in 1980. None of the projects were successful on the first ballot, and all were submitted for a second ballot to all DIPPR® participants. The Data Compilation project was then fully funded, and three other projects received adequate funding to begin: The Data Book Project with help from NBS, and the Acetic Acid and Mixture Properties projects with help from the Environmental Protection Agency. The Electrolyte Phase Equilibria project failed to get enough support in 1979, but was revised and succeeded in 1980 to begin in 1981 at NBS.
The following were elected at the organizational meeting: Al Larsen, Chairman, Data Compilation Project Steering Committee; Bill Seaton (Tennessee Eastman), Chairman, Pure Component Data Project Steering Committee; David Zudkevitch (Allied Chemical), Chairman, Mixture Data Project Steering Committee; Herbert Barner (Kennecott Copper), Chairman, Electrolyte Equilibria Project Steering Committee; Steve Newman, Chairman, Liaison Committee; and Mac Clarke (Olin Chemicals), Chairman, Publicity and Recruitment Committee. The DIPPR® Administrative Committee was formed from those elected, with John Prados Chairman and David Palmer Vice Chairman, subject to AIChE Council approval. In September 1979, John Prados was succeeded by David Palmer as Chairman of the Technical Committee and by David Roth as Chairman of the Administrative Committee.
During 1979, requests for proposals were prepared and distributed, contractors were selected, and preparations were made to begin project work in 1980. At a meeting of the Data Compilation Steering Committee including the Data Book Subcommittee in Gaithersburg, MD, October 26, 1979, Ron Danner and Tom Daubert at The Pennsylvania State University were selected as principal investigators for the Data Compilation and Data Book projects. Grant Wilson, Wilco Research Company (later Wiltec Research, Inc.), was selected as the principal investigator for the Acetic Acid-Water PVT and VLE Measurements Project by its sponsors, and Grant Wilson and John Oscarson, Brigham Young University, were selected as investigators for separate portions of the Mixture Experimental Data project by its sponsors. By January 1980, DIPPR was in full operation with four projects underway.
Rationale: Much of the strength of DIPPR® has come from the rationale developed during its formation. This rationale includes responsiveness to industry needs, annual accountability for funding and results, and independent projects from which DIPPR® participants can choose.
In the late 1970s, available compilations of data did not satisfy industry needs very well. Data were sparse for many chemicals important to industry. Even for common chemicals, data were sparse for some important properties and temperature ranges. Some compilations had been worked on for decades but were still incomplete. Others made heavy use of estimation methods and left the task to the user to resolve the discrepancies between results calculated by different methods. Most available compilations provided no recourse to source data. Values obtained or computed from separate sources often differed significantly. Process engineers had to locate data sources, evaluate data and estimate missing values.
With increasing use of computers for simulation and design calculations, industry representatives believed that satisfying industry needs was essential and had to be accomplished within a relatively short time frame. Data for high-volume chemicals and chemicals important in several processes or in processes run by several companies were needed, and had to include all the properties over all the ranges required for process engineering. Data evaluation would be needed to select the best data and estimate reliable values. Data specialists in sponsor companies would need recourse to source data for data review and further evaluations.
The Data Compilation project became the "flagship" project for DIPPR®, with other experimental projects contributing to the Data Compilation as appropriate. In turn, the Data Compilation would identify gaps and inconsistencies in the literature requiring measurements, which would be prioritized and carried out by other DIPPR® projects.
The DIPPR® balloting procedure gave a sense of urgency for each project to make progress and deliver useful results every year or face the risk of not receiving adequate funding in the next ballot cycle. Thus as projects fulfilled their objectives or were no longer considered useful, they would be terminated by decision of the sponsors or by the balloting procedure. Weighting factors from 1 to 8 gave small companies the opportunity to sponsor DIPPR® projects at low cost. All project results would be published or made available eventually, but sponsors would enjoy exclusive use of the results for at least a year. A summary of DIPPR® projects was recently published in the Journal of Chemical and Engineering Data (1996, 41, 930-934), and illustrates the breadth of scope of DIPPR® projects over the years.
The independence of DIPPR® projects from each other assured that DIPPR® was working on the right things. DIPPR® participants could propose new projects at any time, to respond to new technology needs. Some projects would be of short duration, one to three years, while others would continue longer or indefinitely. Flexibility and relevance would thus be maintained.
Retrospective: This rationale, developed during the formation of DIPPR®, has served remarkably well to provide purpose, stability, and significant results. It continues to be important in guiding DIPPR®. But even more important are the people who have done the work to make DIPPR® successful over the years. This includes Bob Reid, Don Vredeveld and Howard White, whose early concepts and efforts led to the formation of DIPPR®; the Management and Technical Representatives of DIPPR® participants, all of whom have served as volunteers; the DIPPR® Technical Directors, Ted Selover and George Thomson, who have guided the DIPPR® effort with enthusiasm; the AIChE staff, whose support has been crucial; and a large number of contractors, investigators, and students, who have carried out the project work.